US2980532A

US2980532A - Method of producing permeable membranes

Info

Publication number: US2980532A
Application number: US786436A
Authority: US
Inventors: Martensson Karl Marten; Lofman Claes Peter; Eriksson Erik Ingemar
Original assignee: Studsvik Energiteknik AB
Current assignee: Studsvik Energiteknik AB
Priority date: 1958-01-10
Filing date: 1959-01-12
Publication date: 1961-04-18
Anticipated expiration: 1978-04-18
Also published as: GB860099A

Description

A ril 18, 1961 K. M. MARTENSSON ETAL 2,980,532

METHQD OF PRODUCING PERMEABLE MEMBRANES Filed Jan. 12, 1959 .(2) Th y PRODUCING PERMEABLE MEMBRANES Karl Marten Martensson, Claes Peter Liifman, and Erik Ingemar Eriksson, Stockholm, Sweden, assignors to Aktiebolaget'Atomenergi, Stockholm, Sweden, a company Y of Sweden Y I ME OD OF 1 I Thepresent invention permeablem'embranes. These membranes are primarily intended for use in the separation of isotopes by means vof,gasfdiffusion orin thesepar'ation of gas mixtures by means of the same method. A veryimpo'r'tant field of application is the separation of the uranium isotopes.

Membranes to be used for separation by gas diffusion, especially; of the uranium isotopes,- should comply with th'elfollowing requirements: V

(l) The average pore radius should be within the'range of ,100-300 All. .(the requirement of optimum pore Jradius). I I

permeability should be as great as possible (thereqhirement of high'permeability); 1

(3 The membrane should 'possessa good mechanical jstrength 'and be able to withstand pressure differences and vibrations the requirement of strength) 11(4) The membranes should possess agood-chemical resistanceja'gjainst the workingmedium, for instance uranium hexafluoride (the requirement-of corrosion resista s);

I InTmembranes produced by pressing and sinteringof powder thsreis a relation between the average poreradius iand'the average particle size of the' powder; Experience shows that in order to obtain anavera'ge poreradius which is, most favourable for separation of the uranium isotopes, v the diameter of the powder particles should be of theor- 5 der of 1000 AU. a V

It'is obvious that'the above mentioned requirements Zand ,3 are at'variance. The-fact is that it is always possibleto: increase the permeability by decreasing the thickness of the membrane-but generally this cannot be 'done without" neglecting the requirement or: strength. ;When itlcomes to producing plane, sintered membranes of such a'size (about 10 dm. or larger) that'is desirable for a 's'eparationplant on an industrial scale it. is" generally relatespto a method of producing,

necessary, by reason of strength, to make the membranes sothick that considerable concessions as regards permeabilitywill be the result. Inmanufacturing plane, sintered membranes there must be a preparatory working moment-of distributing the powder in'one way or another Lina plane layer; which i's thereupon pressed and sintered. 1 In doing so one must pursue a distribution ina very even jlayerfas otherwise a' pressure'varying overth'e surface of the mfimbrarieis obtained at the pressing, which results in finhomogeneousj membrane' properties, Simi .larly, if :the powder is' not evenly distributed such large pores fnay be formed in the finished membrane that the membrane will beiunfit for git's p irposel' It 'ishobvious that the thinnermembranes one seeks to attain the stronger the requirement must be ofthe powder to be distributed in an even layer from ,the 'beginning, inother words, the

' .thardcrrit will be to avoid inhomogeneous membrane properties; and the occurrence. ofv such membrane becomes unusable. t

large-pores, that the the formationof nitride powder proceeds.

7. 7 The crusts too consist to the greater part of aluminum uranium isotopes.

ice

The present invention relates to a method of producing membranes which are at the same time thin and strong. This method is characterized in that the solid powder particles from which'the membrane is to be produced are suspended in a suspension medium and that the suspension then is brought into contact with a close meshed network in such a way that the particles deposit in a layer on one or both sides of the network which is thereupon pressed and sintered together with the powder layer into a membrane;

In the, following a number of methods will be described of producing powder suspensions having suitable properties. In these methods the suspension medium consists of a gas. s

Trial has thus shown that if aluminium is disintegrated in an electric are Where one or both electrodes consist of aluminium, and if the disintegration takes place in nitrogen, a powder is obtained which will appear in the form of a suspension in the nitrogen immediately after it hasbeenformed. Thus the suspension is fo'rmed automatically in direct connection with the production of the powder. Furthermore, the powder acquires such properties so as to make it particularly adaptable for producing permeable membranes, especially for separation of the The method will be illustrated in greater detail by way 'of an example with reference to the attached drawing showing an apparatus for producing membranes having an area'of 4 dm. "In a receptacle 1 there are arranged two aluminium electrodes 2 and 3,fwhich are fed from a cur-' rent source, not shown. A flow of nitrogen having room temperature and atmospheric pressure is conducted partly deposits as an even coating on the walls of the "receptacle, partly 'entrains in the flow of nitrogen leaving the receptacle. Electron microscopic photographshave shown that the powder consists of an agglomerate of mostly spherical particles having a diameter 'of about 1000 AU. Furthermore, quantitative analysis'has shown that of the powder produced under the conditions above 'S ated consists of aluminium nitride and the rest of metal- -lic'aluminium. I a

For the time being it is uncertain whether the'powder is formed by so called cathodic disintegration,i.e. as a result of the cathode beingbombarded by the positive'ions which areformed in the are, or by a regular volatilization of the electrode material. Possibly both of the said proc- .esses are of importance" for the powder formation. A di rect observation of thecourseof events shows that primarily the cathode is consumed during the process, but -0n the other hand slag-like outgrowths or crusts are formed on the cathode which tend to grow in the direction towards the anode. A"continuous powder formation takes place when the cathode is consumed at the same rate as these crustsmanage'to grow out so that the length proper. of

the ,arc remains constant. Such an equilibriumjsdifiicult to carry into effect, and as a rule the crusts grow more On the other'hand the-crusts-are different portions 'of the crusts. As long as this takes place from the art of electrical filters. r h After the powder layer (or powder layers) hasreached 3 nitride, and it seems plausible that they originate by disintegration at the cathode of aluminium which immediately combines with the nitrogen to form aluminium nitride, which in turn wholly or partially deposits in the form of a coating on the cathode. Whether the aluminium nitride appearing in the form of a powder in the gas phase likewise is formed by a cathode disintegration in a proper sense, or secondarily by a volatilization of the said crustlike coating, has not been .possible to settle for the time .being.

7 After the arc has goneout in consequence of the fact that the nitride crust emanating from thecathode has grown outso that the gap between the electrodes. is

bridged, it may be relighted after the crust has been removed. This may usually be done from the outside by the electrodes, which are slightly movable laterally, being rubbed against one another so that the crust loosens.

The figure also shows a device for depositing the powder 'ona metallic wire network. The device according to the figure is provided with a lid 4 which may be screwed on by means of eight finger nuts 5. The plate receptacle is providedwith two gas inlets 6 connected to a supply pipe 16, and the lid is provided with a gas outlet 7 having a valve 8. The two metal electrodes 2 and 3 are held in position by means of Teflon bushings 9. Alternatively the receptacle may be provided with two or more pairs of electrodes. A network 10 is secured in a mounting frame .11, which in turn is clamped between the lid and the receptacle. 12 and 13 are rubber gaskets which partly ensure that the clamping becomes tight and partly insulate the network from the receptacle. In the wall of the receptable two diametrically opposed glass windows are mounted so that it becomes possible to control the are from the outside. When the gas flow with suspended particles of the powder formed in the are passes the network, the particles arecaught and form an even deposit on the wire network. When the deposit becomes'thicker its flow resistance increases and a pressure is created in the receptacle. This pressure is opposed, however, by the overpressure valve 14, which consists of a branch pipe 15 provided in the supply conduit and extending down into a containerwith mercury. Q

In a number of trials. nickel networks have been used having a wire diameter of 35 microns and a meshwidth of 40 microns. On such a network a particularly even deposition of powder is obtained. The powder may be applied one side only or on both sides, in which later case the mounting frame 11 is turned after a layer has been 'formed on one side of the network. r

It is possible to improve still thehomogeneity of'the powder deposit obtained on the wire network in. the man- -ner stated by an electric voltage being appliedto the network, which voltage essentially deviates from 'the voltages of the two electrodes. Alternatively, one or more metal wires may be provided adjacent the network and an electric voltage be applied between the network and these aesomsa A .t

strength. If the above stated trial conditions prevail and if the compression pressure is 400 kp./cm. the sintering temperature 750 C. and the sintering period 5 min., membranes are obtained having permeabilities in the range of 1.2-2.5 cm. air at N.T.P. per sec. and cm. membrane area at a pressure difference ofl atm. and average pore radii in therange of 175-350, A.U.

Under the conditions stated above concerningthe voltage and current of the are, a powder was obtained to :about 90% consisting of aluminium nitride. Trials have shown, however, that if the current is raised to considerably higher values a powder is obtained which contains a larger share of metallic aluminium. It is thus possible to produce a-powderhaving a content of. 40% aluminium, if a current of about 75 amperes is used. The ground for the higher aluminium content in this case seems to be that the aluminium of the electrodes volatilizes and then immediately condenses, there being no time for any nitride formation to take place.

For the time being it is uncertainwhether membranes produced of a powder to the greater part consisting of aluminium nitride are sufficiently resistant against UF The trials hitherto carried out, with these membranes in order to establish the corrosion in uranium hexafluoride have not given unequivocal results. The fact, however, that aluminium nitride is comparatively very resistant against halogens, which has been proved, indicates that aluminium nitride also should possess a high resistance against uranium, hexafiuoride. i

It is known that aluminium oxide is eminently resistant .against uranium hcxafiuoride. Trials have shown that if a powder, which to the greater .part consistsof aluminium nitride and which is producedin the manner stated above, is treated withair or-oxygen. at an elevated temperature, the aluminium nitridewill be transformed partly. into aluminium oxide. At one trial the powder wasvthus heated to 750 C. for 16 hours anditwas established by analysis that 35% of the powder hadbeentransformed into oxide. The colour of the powder is changed by this treatment from grey to white. -It is likely enoughthat during the oxidation a surface layer of aluminium oxide isformed which surrounds the powder particles and protects the inner portions, against attack, for which "reason a complete oxidation is ,difficultrto carry out. i r

.Membranes produced in the manner'stated above ofa powder to the greater part consisting of aluminum nitride may also be transformed partlyinto aluminum oxide by being heated in air or oxygen. Thus in anumber of trials membranes wereheated to 750 C. for 4 hours and it was proved that. ;on the average 25%, of the nitride had changed into oxide. The colour of the-membranes was wires. In both cases an effect is obtained which is known sufiicient thickness the arc is broken and the nitrogen new cut off. The wire network is-loosened from the mounting frame 11 and is compressed together with the powder be-. tween the plane surfaces of a pressing tool. Compression pressures varying between 1000 and 5000 kp./cm. have been used. At the compression apart of the powder-is pressed into the meshes of the network and a very strong combination is obtained. Thismere compression renders the membrane so strong that itfm ay be readily handled.

However t'he strength is still more improved by sintering the membrane at an elevated temperature. Sinter'ing terriperaturesvarying between 600 and 800 C. and-sintering periods between 3 and 60 minutes have been used! The sintering has i.a. been performed in a hydrogen at mo s- H phere. Thefinished membranes have a thickness-of about 50 microns only, but nevertheless they possess a very high changed from grey to white. Itislikely in this case too that a protective. layer is, formed on those surfaces of the membrane which are incontact with the air or oxygen. By causing the air or oxygen to flow through the membrane during the heating it is possible to attainthe feature .thatthe walls of. the pores in theinterior of the membrane are coated with alayer of aluminum oxide. Thus, if membranes made of a powder of aluminum nitride are insufficiently resistant against uranium. hexafluoride it is possible, by treating the membranes in the abovestated way, to-coat thoseinterior surfaces which are in contact with uranium hexafluoride with a protective layer of aluminum oxide, whereby a perfectly satisfactory resistance against uranium. hexafluo'ride is attained.

, It has been found that the strength of the membranes is not impaired bythe treatment stated and that their separating capacities and permeabilities remain substantially unchanged. I i

A membrane inwhichthe aluminum nitride has been partly transformed into aluminum oxide, may alsp' be preparedby sintering the membrane in airor oxygen,

' duced under the above stated conditions.

between two foils of a metal having a great hardness after which the network is rolled together with the two foils between cylindrical rollers and at a suitable pressure of rolling. Thereupon the membrane is sintered as stated above. Trials have shown that membranes rolled and sintered in this way possess qualities which are entirely comparable with those of membranes which have been pressed instead of rolled and then sintered. In producing large membranes rolling may be technically simpler,

' quicker and cheaper than pressing, and it may permit a continuous method of production. The rolling may also be performed without the wire network with the powder layer first having to be placed between two metal foils.

At the disintegration or volatilization of aluminum in an arc in the manner described above, the nitrogen may wholly or partially be replaced by air or oxygen (or with some other gas mixture containing oxygen and an inert gas). In this case a powder will be formed which consists of aluminum oxide provided that there is a sufiicient percentage of oxygen in the gas mixture. This powder appears in the beginning in suspended form in the gas phase and may be deposited on a close meshed network in the same way as described above. Trials have thus been made by means of the apparatus shown in the figure, a gas mixture'being used having a content of 5-10% oxygen and nitrogen for the rest. The voltage between the aluminum electrodes was about 30 volts and the current was about 7 amperes. A powder was obtained consisting of pure aluminum oxide. Membranes made of this powder and for the rest in the manner described above (pressing at 4000 kp./cm. sintering for 5-30 minutes at 110 C.) have shown permeabilities in the range of 1.2-2.0 cm. air at N.T.P. per sec. and cm. membrane area at a pressure difference of 1 atm., and average pore radii in the range of 300-450 A.U. At insufiicient supply of oxygen to the receptacle a powder is formed which consists of aluminum oxide and aluminum nitride in a quantity ratio determined by the local oxygen content around the arc.

sintering. Inthe said ways membranes havebeen produced having permeabilities in the range, of 0.2-1.0 cm. air at N.T.P. per sec. and cm. membrane area at-a pressure diiference of 1 atm., and average pore radii in the range of 300-800 A.U.

r The three methods described above for producing a powder for permeable membranes in such a way that the powder is suspended in a gas phase in connection with its production and thereafter is deposited on a close meshed network, are all based upon disintegration (or volatilization) in an electric are. It is obvious that this arc method offers extensive possibilities of variation.

Thus other metals than those mentioned here are conceivable as electrode material. Further, other gases (or gas mixtures) than those occurring above may be selected. Furthermore, the pressure and temperature of the gas may be varied. Finally it may be possible to change the Trials have also been made in which nickel electrodes have been disintegrated or volatilized in an arc in nitrogen of atmospheric pressure and room temperature. The

voltage across the electrodes was about 30 volts and the v current was about 10 amperes. -A powder was obtained tions are necessary, however, in producing membranes of nickel powder inasmuch as the sintering of the powder layers deposited on a close meshed network must be .per-

formed under pressure, as otherwise microscopical holes are formed as a consequence of the contraction during the sintering, which holes make the membrane unusable. This contraction, which as a rule occurs at the sintering of nickel membranes, is efficiently counteracted by the sintering taking place under pressure. In a number of trials membranes have been made of nickel powder pro- The networks with the deposited powder layers were first pressed at room temperature at a pressure of 1000-4000 kp./cm. and'were then sintered at 750-850 C., the compression pressure being 1000-2000 kp./cm. In other trials the precompression was omitte'd and' only one compression 7 v operation was made, thatis' in connection with the 75 -10. A method'for the production of a permeable memproperties of the powder by varying the voltage and current oil the arc.

The possibilities of producing a powder-for permeable membranes in such a way that the powder is suspended in a gas phase in connection with its production and thereupon is deposited upon a close meshed network according to the present invention are not restricted, however, to the arc method described above. For instance, it is possible to burn, in a manner known per se, volatilized metal salts in air or oxygen (or oxygen mixtures) so as to form the corresponding metal oxide which may be obtained in the form of a powder which occurs in suspended form in agas phase and is then deposited, in accordance with the present invention, on a close meshed network and compressed and sintered to fonn permeable membranes. Thus, volatilized aluminium chloride may be burnt in an atmosphere containing oxygen, so as to form a powder of aluminium oxide.

In the foregoing methods have been described according to which the suspension medium has been a gas and the suspended particles have been metals or metal compounds. As a matter of course the method according to the invention may also be carried into effect with liquid suspension media. Likewise the suspended particles may consist of other materials than metals or metal combinations, for instance, organic materials such as plastics. As regards membranes for separation of the uranium isotopes the carbon fluorides, such as those of the Teflon type, which are resistant against uranium hexafiuoride, are of special interest.

What is claimed is:

1. A method of producing a permeable membrane which comprises vaporizing a metallic electrode in a gas atmosphere by means of an electric are, collecting finely divided material from said atmosphere directly on a metallic network, compressing said collected material together with the network to form a compressed unit and sintering the resulting compressed unit.

2. A method as defined in claim 1 in which the gas atmosphere is caused to flow through said network.

3. A method as defined in claim I in which an electric potential is impressed upon said network to attract said finely divided material thereto. I

4. A method as defined in claim 1 in which the electrode is aluminum and said. atmosphere is nitrogen.

5. A method as defined in claim 1 in which the electrode is aluminum and said atmosphere contains oxygen.

6. A method as defined in claim 1 in which the electrode is nickel.

7. A'method as defined in claim 1 in which the compression is effected by rolling. r

8. A method as defined in claim 1 in which the metal is aluminum and the compressed and sintered unit is heated in an atmosphere containing oxygen.

9. A method as defined in claim 1 in which the sintering is effected in an atmosphere containing oxygen.

brane which compfisesburning-a frietal'lsalf in an atmo'spliers-containing, oxygen, depdsiting fi'nelydiyided material from said atmosphere onuinetallic ngtyyorhtompressingthe deposit and-sinter-ing1he compressed unit together with the network to form a compressed unit;

11'. A method as defined, in claim :10 in which the.

metal salt is an aluminum salt.

1 I Referenc'es ('JiterL-iu thefile ofgthisqpatent 'UNITED STATESPATENTS Morf Feb.j9, 1915 Powell"; ;;Scpt. 26, 1933 Wissler Feb. 6, 1940 Lepsoe Mar. 6, 1945 UNITED STATES PATENT eTTIeE CERTIFICATIQN @F CORRECTION Patent No. 2 980532 April 18 1961 9 Karl Mar-ten Martensson eta 5110 It is hereby certified that error eppears in the above numbered patent requiring correction and that the said Letters Patent should read as corrected below.

Column 7 lines 4 and 5 strike out together with the network to form a compressed uni-T, and insert the same after "deposit" in line 4 same column '3. 1

j Signed and sealed this 22nd day of August 1961.

(SEAL) Attest:

ERNEST W. SWIDER Attesting Officer DAVID L. LADD Commissioner of Patents UNITED STATES PATENT OFFICE CERTIFICATION OF CORECTION Patent No. 2 980582 April 18 1961 Karl M arten Martensson et 31.

It is hereby certified that error sppears in the above numbered patent requiring correction and that the said Letters Patent should read as corrected below.

Column 7, lines 4 and 5, strike out togetl'ler with the network to form a compressed unit? and insert the same after "deposit", in line 4 same column 7.

Signed and sealed this 22nd day of August 1961.

(SEAL) Attest:

ERNEST W. SWIDER Attesting Officer DAVID L. LADD Commissioner of Patents

Claims

1. A METHOD OF PRODUCING A PERMEABLE MEMBRANE WHICH COMPRISES VAPORIZING A METALLIC ELECTRODE IN A GAS ATMOSPHERE BY MEANS OF AN ELECTRIC ARC, COLLECTING FINELY DIVIDED MATERIAL FROM SAID ATMOSPHERE DIRECTLY ON A METALLIC NETWORK, COMPRESSING SAID COLLECTED MATERIAL TOGETHER WITH THE NETWORK TO FORM A COMPRESSED UNIT AND SINTERING THE RESULTING COMPRESSED UNIT.